PIE in the Sky : Online Passive Interference Estimation for Enterprise WLANs

Transcription

1 WiNGS Labs PIE in the Sky : Online Passive Interference Estimation for Enterprise WLANs * Nokia Research Center, Palo Alto Shravan Rayanchu, Suman Banerjee University of Wisconsin-Madison Konstantina Papagiannaki Intel Labs, Pittsburgh *vivek.2.shrivastava@nokia.com NSDI

2 Enterprise WLAN setting Clients Internet Access Point Wireless controller NSDI

3 Enterprise WLAN setting Clients Functionalities implemented at controller Intrusion detection system Interference management Internet (channel assignment, power control) Access Point Wireless controller NSDI

4 Enterprise WLAN setting Clients Functionalities implemented at controller Intrusion detection system Interference management Internet (channel assignment, power control) Access Point Wireless controller NSDI

5 Problems with wireless* The wireless is being flaky. Flaky how? Well my connection dropped earlier and now it seems to be slow We will take a look. User Wait, now it seems fine. Support *Slide borrowed from Cheng et. al (Jigsaw, Sigcomm 06) NSDI

6 Problems with wireless Hidden terminals 6 6 NSDI

7 Problems with wireless Rate anomaly 6 Mbps 54 Mbps Mismatch in data rates, slows down fast links NSDI

8 Problems with wireless Increasing client density and mobility. 8 8 NSDI

9 Problems with wireless Increasing client density and mobility. 9 9 NSDI

10 Problems with wireless. changing interference patterns NSDI

11 Interference management in WLANs NSDI

12 Interference management in WLANs Estimate Interference dynamically Manage Interference (data scheduling, transmit power control, channel assignment) NSDI

13 Interference management in WLANs Estimate Interference dynamically Manage Interference (data scheduling, transmit power control, channel assignment) NSDI

14 How to estimate interference? Use bandwidth tests NSDI

15 How to estimate interference? AP-Client Use bandwidth tests pair Interferer NSDI

16 How to estimate interference? 1) Measure AP-Client delivery in isolation NSDI

17 How to estimate interference? 1) Measure AP-Client Isolation delivery in isolation delivery= 0.95 NSDI

18 How to estimate interference? 1) Measure AP-Client delivery in isolation 2) Activate interferer and measure delivery NSDI

19 How to estimate interference? 1) Measure AP-Client Interference delivery= 0.66 delivery in isolation 2) Activate interferer and measure delivery NSDI

20 How to estimate interference? 1) Measure AP-Client delivery in isolation 2) Activate interferer and measure delivery Link Interference Ratio (LIR) = del Interference / del isolation NSDI

21 How to estimate interference? 1) Measure AP-Client delivery in isolation 2) Activate interferer and measure delivery Link Interference Ratio (LIR) = 0.66 / 0.95 = 0.69 NSDI

22 How to estimate interference? 1) Measure AP-Client delivery in isolation 2) Activate interferer and measure delivery 0 LIR 1 Strong Weak NSDI

23 But are bandwidth tests practical? Can we use bandwidth tests in live settings Good accuracy Network downtime required - X Not scalable (~ 1 hr for 20 AP-Client pair network) - X Not based on realistic rates and packet sizes X Inefficient in dynamic scenario (client mobility) X NSDI

24 But are bandwidth tests practical? Can we use bandwidth tests in live settings Good accuracy Network downtime required - X Not scalable (~ 1 hr for 20 AP-Client pair network) - X Not based on realistic rates and packet sizes X Inefficient in dynamic scenario (client mobility) X Can we estimate interference in a passive, real-time way? NSDI

25 PIE Outline Motivation Conventional bandwidth tests not sufficient Passive Interference Estimation (PIE) Polling period of PIE Accuracy of PIE Realistic trace replay with PIE Applications of PIE Summary NSDI

26 Estimating interference passively Sniffer Sniffer NSDI

27 Estimating interference passively Sniffer could be a dedicated wireless radio Clocks synchronized using wired backplane Sniffer Sniffer NSDI

28 Estimating interference passively Sniffer reports NSDI

29 Estimating interference passively Sniffer reports Timestamp, duration, rate, success.. NSDI

30 Estimating interference passively Hidden terminals NSDI

31 Estimating interference passively Hidden terminals NSDI

32 Estimating interference passively Hidden terminals NSDI

33 Estimating interference passively Hidden terminals 1. Carrier sense NSDI

34 Estimating interference passively 1) Note timestamp, Hidden terminals rate, duration 2. Channel free, transmit NSDI

35 Estimating interference passively 1) Note timestamp, rate, duration 2) Note if transmission is a success (ack received?) 3. Collision! NSDI

36 Estimating interference passively Timestamp: T0 Rate: 6Mbps Length: 1400 bytes Success: False Timestamp: T0 + δ Rate: 12Mbps Length: 600 bytes Success: False NSDI

37 Estimating interference passively T0 6Mbps 1400 bytes False T0 + δ 12Mbps 600 bytes False Interference estimation NSDI

38 Estimating interference passively T0 6Mbps 1400 bytes False T0 + δ 12Mbps 600 bytes False NSDI

39 Estimating interference passively T0 6Mbps 1400 bytes False T0 + δ 12Mbps 600 bytes False NSDI

40 Estimating interference passively T0 6Mbps 1400 bytes False T0 + δ 12Mbps 600 bytes False & NSDI

41 Estimating interference passively Red & T0 6Mbps 1400 bytes False T0 + δ 12Mbps 600 bytes False & Green clients interfere NSDI

42 Estimating interference passively Infer interference Sniffer reports cenarios eception X X X X NSDI

43 Estimating interference passively Infer interference Sniffer reports cenarios eception X X X X Red and Green packets overlaps => both lost NSDI

44 Estimating interference passively Infer interference Sniffer reports cenarios eception X X X X No overlap, no problem! NSDI

45 Estimating interference passively Infer interference Sniffer reports cenarios eception X X X X Both way hidden terminals NSDI

46 Estimating interference passively Infer interference Sniffer reports cenarios eception X X NSDI

47 Estimating interference passively Infer interference Sniffer reports cenarios eception X X Red and Green packets overlaps => Green is lost 47

48 Estimating interference passively Infer interference Sniffer reports cenarios eception X X One way hidden terminals NSDI

49 Computing interference measure in PIE Compute Isolation loss rate Fraction of non-overlapping packets lost Compute Interference loss rate Fraction of overlapping packets lost Interference measure (LIR): (1 Interference loss) / (1 Isolation loss ) NSDI

50 How quickly can PIE converge? Time taken by PIE to converge depends on two key properties Periodicity with which sniffer reports are collected by the controller Traffic patterns for the links which dictate the number of interference events captured in a time interval NSDI

51 How quickly can PIE converge? Time taken by PIE to converge depends on two key properties Periodicity with which sniffer reports are collected by the controller What is the minimum polling period? Traffic patterns for the links which dictate the number of interference events captured in a time interval How much time does PIE take under realistic access patterns? NSDI

52 PIE Outline Motivation Conventional bandwidth tests not sufficient Passive Interference Estimation (PIE) Polling period of PIE Accuracy of PIE Realistic trace replay with PIE Applications of PIE Summary NSDI

53 What is the minimum polling period? P P P time interval I 1 I 2 Time NSDI

54 What is the minimum polling period? P P P time interval I 1 I 2 Time R(I 1 ) R(I 1 ) NSDI

55 What is the minimum polling period? P P P time interval I 1 I 2 Time R(I 1 ) R(I 1 ) LIR (I 1 ) NSDI

56 What is the minimum polling period? P P P time interval I 1 I 2 Time R(I 2 ) R(I 2 ) LIR (I 1 ) NSDI

57 What is the minimum polling period? P P P time interval I 1 I 2 Time LIR (I 1 ) R(I 2 ) R(I 2 ) LIR (I 2 ). NSDI

58 LIR What is the minimum polling period? Polling period (ms) Stability of interference measure for saturated traffic NSDI

59 LIR What is the minimum polling period? Measure stabilizes after ~85 ms (at least 20 overlap samples) Polling period (ms) Stability of interference measure per polling period NSDI

60 LIR What is the minimum polling period? We use a polling period of 100ms Polling period (ms) Stability of interference measure per polling period NSDI

61 PIE Outline Motivation Conventional bandwidth tests not sufficient Passive Interference Estimation (PIE) Polling period of PIE Accuracy of PIE Realistic trace replay with PIE Applications of PIE Summary NSDI

62 % of link-interferer pairs How accurate is PIE? Mean Error in LIR estimation 62 62

63 % of link-interferer pairs How accurate is PIE? Mean Error in LIR estimation 63 63

64 % of link-interferer pairs How accurate is PIE? 95% of link-interferer pairs, LIR computed by PIE is within +/- 0.1 of the value reported by BW test Mean Error in LIR estimation 64 64

65 PIE Outline Motivation Conventional bandwidth tests not sufficient Passive Interference Estimation (PIE) Polling period of PIE Accuracy of PIE Realistic trace replay with PIE Applications of PIE Summary NSDI

66 PIE with realistic access patterns NSDI

67 PIE with realistic access patterns Evaluate PIE using realist traffic patterns on a 15 node topology (7 AP 8 laptops) Each client laptop replays the traffic patterns of an actual client from a real wireless trace Three activity periods: heavy (> 40 % medium busy), medium (40 20% busy), light (< 20% busy) NSDI

68 Time to estimate (ms) PIE with realistic access patterns Heavy Medium Light Traffic period NSDI

69 Time to estimate (ms) PIE with realistic access patterns Convergence is faster for higher client activity 0 Even for light activity, Heavy median Medium time of estimate Light LIR is less than 650 ms Traffic period NSDI

70 PIE Outline Motivation Conventional bandwidth tests not sufficient Passive Interference Estimation (PIE) Polling period of PIE Accuracy of PIE Realistic trace replay with PIE Applications of PIE Summary NSDI

71 What is the impact on WLAN applications? AP-Client pairs NSDI

72 What is the impact on WLAN applications? AP-Client pairs Evaluate usefulness of PIE for an interference mitigation mechanism (data scheduling using CENTAUR Mobicom 09) NSDI

73 What is the impact on WLAN applications? AP-Client pairs 1. Estimate interference using PIE NSDI

74 What is the impact on WLAN applications? AP-Client pairs 1. Estimate interference using PIE 2. Input estimate to a centralized data scheduler NSDI

75 What is the impact on WLAN applications? AP-Client pairs 1. Estimate interference using PIE 2. Input estimate to a centralized data scheduler 3. Evaluate performance under dynamic scenarios NSDI

76 System Throughput (Mbps) What is the impact on end users? Static Static scenario NSDI

77 System Throughput (Mbps) What is the impact on end users? Static Static scenario NSDI

78 System Throughput (Mbps) What is the impact on end users? Static scenarios, PIE is comparable to BW test Static Static scenario NSDI

79 System Throughput (Mbps) What is the impact on end users? Mobile Mobile scenario NSDI

80 System Throughput (Mbps) What is the impact on end users? Mobile Mobile scenario NSDI

81 System Throughput (Mbps) What is the impact on end users? Mobility scenarios, PIE outperforms BW test Mobile Mobile scenario NSDI

82 What is the impact on end users? PIE can also be used to monitor production systems (like Jigsaw) We monitored two production WLANs Use testbed nodes in proximity of production APs as sniffers Identify hidden terminals and rate anomaly problems NSDI

83 What is the impact on end users? WLANs Hidden terminal cases (LIR < 0.7) Rate anomaly cases (Ratio of rates < 0.2) WLAN1 8% 21% WLAN2 11% 22% NSDI

84 What is the impact on end users? WLANs Hidden terminal cases (LIR < 0.7) Rate anomaly cases (Ratio of rates < 0.2) WLAN1 8% 21% WLAN2 11% 22% Hidden terminals are rare, but can become pain points for clients Rate anomaly is more frequent, but do not cause drastic performance issues NSDI

85 PIE Outline Motivation Conventional bandwidth tests not sufficient Passive Interference Estimation (PIE) Polling period of PIE Accuracy of PIE Realistic trace replay with PIE Applications of PIE Summary NSDI

86 Related Work PIE leverages techniques from Jigsaw, WIT (Sigcomm 2006) and builds on their ideas Focus of Jigsaw, WIT was to understand interference, ours is to compute it in real-time CMAP also infers interference to harness exposed terminals, but requires physical layer change Active techniques like Microprobing (CoNext 2008) still require downtime and do not use realistic traffic NSDI

87 PIE Limitations Does not handle non-wifi interferer like microwaves. Can miss external interferers if none of the enterprise APs can listen to the interferer May miss client conflicts, can use client participation in PIE to enhance the system Interference detection techniques at the physical layer may be more accurate in some scenarios where diversity is too low for PIE to function NSDI

88 PIE Summary Online interference estimation important for interference mitigation BW test incurs high overhead, requires downtime PIE is a passive mechanism, generates interference estimates in real time Leverages centralized infrastructure to collect real time reports from APs Non-intrusive with good accuracy NSDI

89 Thank you! NSDI